FDA commissioner Stephen Hahn speaking at a White House press conference on 23 August. He later clarified his remarks in a tweet
SAUL LOEB/AFP via Getty Images
Blood plasma donated by people who have recovered from covid-19 will be used as a treatment for the infection in the US. The US Food and Drug Administration (FDA) granted an emergency use authorisation for the treatment on 23 August, but the evidence that it works is lacking.
Convalescent plasma is known to have been used to treat pandemic flu back in 1918. It involves collecting blood plasma – the yellow liquid component of blood stripped of its blood cells – from people who have recovered from a disease. The plasma can contain antibodies generated by the immune system to fight or prevent a future infection, although the antibody levels vary between donors.
The treatment does appear to work for some infections, such as diphtheria, but research has been spotty, and there has been a lack of randomised, placebo-controlled trials, says Lise Estcourt at the University of Oxford. More recently, the treatment was found to be ineffective for Ebola.
Several studies are under way to test convalescent plasma for covid-19. The largest has been run by the Mayo Clinic in the US – about 71,000 people have received treatment across 2780 hospitals over the past five months as part of a programme that enables access to experimental therapies.
Based on the data collected from around 35,000 of these individuals, the researchers behind the project found that people treated with plasma containing higher levels of antibody, and those treated earlier in the course of their illness, appear less likely to die within a seven or 30-day window.
But because the study wasn’t randomised, and none of the participants received a placebo, it is impossible to draw firm conclusions from the data. It isn’t clear if some other factor might have improved survival rates in some people. The treatment of and survival from covid-19 has improved over time, for example.
“We’re still not certain of the effectiveness,” says Estcourt, who is part of a team running a clinical trial of convalescent plasma in the UK.
The lack of evidence has led several high-profile scientists and doctors – including senior US health officials Francis Collins and Anthony Fauci, according to reports from The New York Times – to caution against FDA authorisation.
Despite such warnings, the emergency use authorisation was announced on 23 August in a White House press briefing by US president Donald Trump and Stephen Hahn, the head of the FDA. During the briefing, Hahn stated that, for the sickest individuals given the optimum treatment, “(for) 100 people who are sick with covid-19, 35 would have been saved because of the administration of convalescent plasma”.
But this statistic isn’t supported by data, say researchers contacted by New Scientist. “It was unhelpful (for Hahn) to phrase it that way,” says Estcourt. The FDA’s memorandum detailing the authorisation refers to the Mayo Clinic study. In that paper, which hasn’t yet been peer reviewed, the authors describe a 35 per cent lower relative risk of death in individuals given plasma with high levels of antibody compared with those given infusions with low levels, over a seven-day period.
But this figure represents a difference in the relative risk of death, not the absolute risk. In this case, the 35 per cent reduction represents a difference between 13.7 per cent and 8.9 per cent absolute risk. There is no published evidence to support the claim that convalescent plasma can reduce the risk of death by 35 per cent. “From what I’ve seen, people have said it’s more like 3.5 per cent,” says Stephen Griffin at the University of Leeds, UK.
The information that Hahn presented was clearly wrong, says Ashish Jha at the Harvard T.H. Chan School of Public Health. “It’s really quite disturbing.”
The FDA didn’t respond to New Scientist’s request for comment, but Hahn has since tweeted to clarify his remarks. “What I should have said better is that the data show a relative risk reduction not an absolute risk reduction,” he said.
Fast-tracking the use of a therapy without evidence to support its use could “waste an opportunity” to develop a potentially useful treatment, says Griffin. Further research might reveal, for example, that convalescent plasma only works for some people, in certain stages of disease, and only from a proportion of donors, he says. “It could be very promising, but if we rush it, we could not give it appropriately,” he says. “Or it might not actually work.”
The FDA’s authorisation of hydroxychloroquine was based on similar research findings, says Griffin. The drug, which had been regularly touted by Trump, was issued an emergency use authorisation for covid-19 on 28 March. But after growing evidence that the treatment wasn’t effective, and could have serious side effects, the authorisation was revoked on 15 June.
Allowing the use of the therapy could also scupper research efforts, at least in the US, warns Jha. “You can end up with a situation in which people say: ‘I’d rather skip the trial and get the therapy directly’,” he says.
Sign up to our free Health Check newsletter for a round-up of all the health and fitness news you need to know, every Saturday
More on these topics: